Neuropsychological tests have poor specificity among minority populations and thus misdiagnosis of dementia is thus more likely among cognitively normal African Americans as compared to non-Hispanic Whites. During the prior funding period, this study directly addressed the problem of misdiagnosis of cognitive impairment and dementia among African American elders by measuring cultural and educational factors that differ within and between racial groups, and then using these data to understand of the role of race and culture on cognitive test performance. Our work identified aspects of cultural and educational experience that could be explicitly measured and related these variables to test performance both cross-sectionally and longitudinally. We found that after accounting for literacy level (an estimate of quality of education), differences in the test performance of education- matched African American and White elders largely disappeared. Regardless of race, literacy level was the strongest predictor of performance on measures of verbal and nonverbal ability and the best predictor of future memory decline. Despite this wide-ranging effect of literacy on cognition, several researchers have suggested that the primary influence of literacy is on the functioning of a specific buffer or short-term storage system, the phonological loop. However, this theory has been put forth without adequately assessing visuospatial short-term memory or well characterized tasks of working memory. To date, this project has generated major insights into the role of cultural and educational experience on cognitive test performance and the risk of cognitive decline over time. The current proposal will use these insights to guide new studies of the relationship between culture, literacy, and cognition. The three aims of the study are to: 1) use resource characteristics and a broad measure of literacy skill to determine if a measure of reading recognition is a true reflection of quality of education; 2) determine whether school resource variables are the best predictors of cognitive decline and development of dementia regardless of race; and 3) test the hypotheses that literacy affects storage of visuospatial material and the functioning of the central executive or working memory, by administering specific and well characterized tasks of storage and manipulation of verbal and visuospatial information.